Choose the right method for each type of conclusion.

The most common mistake in the brand diagnosis phase is not a lack of data but using the wrong data collection method for the question that needs answering. You survey when you should be interviewing, or trust social media comments as if they represent the entire customer base. The result is having numbers but not being able to make decisions, or making the wrong decisions due to incorrect evidence.

Four methods and questions each can answer

Each method is created to answer a specific type of question. If used incorrectly, data may be available but it won't lead to the necessary conclusions.

• Survey (quantitative research): answers the questions "what percentage" and "what ratio." Use it when you need numbers to compare two points in time, two customer groups, or need data convincing enough for internal or partner use.
• Qualitative interviews: Answering the questions "why" and "what is really going on in the customer's mind." Interviews uncover motivations, attitudes, and beliefs hidden behind behavior, things that surveys cannot reach.
• Buy mystery shopping: Answer the question "how does the actual service experience unfold at each touchpoint?" This is observational research, recording what really happens, independent of what customers remember or choose to recount.
• Social media scanning and platform evaluation: answer the question "what do customers say about the brand when not prompted". Useful for capturing natural language and genuine concerns of customers, but it is important to understand the inherent biases of this channel.

When to choose which direction, when to combine two directions

Common mistakes when selecting a method

• Using one method for all conclusions: often surveys because they are familiar and cheap. The result is a percentage but no understanding of the cause, leading to misguided interventions.
• Leading survey question: ask "Are you satisfied with our professional service?" instead of asking neutrally. It looks better but does not reflect reality.
• Viewing social media as representative of the entire audience: Public writing groups often lean towards extremes, either praising or criticizing strongly. Regular customers rarely speak up.
• Skipping anonymous test purchases to save costs: The consequence is never seeing which touchpoints are genuinely failing. Surveys cannot replace this step because customers do not remember or do not fully articulate what happened.

No method can cover everything on its own. Good practice is to combine testing for service accuracy, surveying for feedback, and interviewing for depth.

Intouch Insight, guide to coordinating research methods

How Sinh Vũ combines methods by layer

In Sinh Vũ's five-layer audit, the method chosen is based on the layer being examined, not simply on habits or budget.

• Brand perception layer: focuses on surveys to measure recognition, associations, and trust levels.
• Touchpoint layer: use real samples and conduct mystery shopping. Sinh Vũ collects between 30 to 50 real touchpoint samples during the collection week, from packaging, invoices, to customer service calls.
• Positioning layer: based on interviews with leaders combined with competitor comparisons, as positioning is a strategic assessment that requires depth, not just ratios.

In the collection week, Sinh Vũ listens to both customers and the internal team to get both sides. The team often describes how touchpoints operate theoretically; customers share what they actually encounter. The discrepancies between the two sides are the real issues that need to be addressed.

You do not need to run all four methods at the same time. Start by clearly defining the question that needs to be answered, then choose the most suitable method for that question. If you need multiple types of conclusions, layer additional methods rather than applying one method to all.